Available online at www.sciencedirect.com
Energy
Procedia
Energy Procedia 4 (2011) 5549–5556
Energy Procedia 00 (2010) 000–000
www.elsevier.com/locate/procedia
www.elsevier.com/locate/XXX
GHGT-10
Towards large scale CCS
Trina Drehera, Craig Dugana1*, Trent Harkinb, Barry Hooperb
a
Process Group, 5 Hobbs Court, Rowville, Vic., 3178, Australia
b Cooperative Research Centre for Greenhouse Gas Technologies (CO2CRC), The University of Melbourne, Vic., 3010, Australia
Elsevier use only: Received date here; revised date here; accepted date here
Abstract
In order to reduce CO2 emissions on a global scale large pilot and demonstration projects that trial new technologies, designs, or
construction techniques applicable to full scale plants need to be undertaken. Process Group has designed and built several pilot
scale capture plants including one located at the Hazelwood Power Station, which is the largest capture plant on a coal fired
power station in Australia. This paper discusses some of the lessons from these pilot plants and presents a new retrofit postcombustion study that investigates carbon capture from a 500MW power station (nominally 3.7 million tpa CO2) using three
solvents with and without heat integration into the steam cycle. Data pertaining to processing 25% of the flue gas from a 500MW
power station (nominally 0.9 million tpa CO2) is also presented.
The study found <5% difference between existing solvent processes in terms of overall plant CAPEX and <15% difference in
OPEX and that for the end user the most advantageous way to design a capture plant is to ensure that it functions with a wide
range of solvents and can be easily adapted for future technology advances. The cost of capture for a 500 MW brown coal power
station with non-optimised heat integration was determined to be in the range AUD$53-63/t CO2 avoided, which incorporated an
improvement of approximately $8-13/t due to the non-optimised heat integration. The heat integration resulted in modest (5%)
energy and cooling duty savings however, with further optimisation performed specific to the given power station and capture
plant it is expected that the cost of capture could be further reduced to at or below AUD$50/t CO2 avoided. In situations where
cooling water is used exclusively for a full scale capture facility the cooling water usage increased by 85-95%. However, when
enhanced heat integration is incorporated this increase is expected to be limited to 75-80% for all technologies analysed.
c© 2011
⃝
by Elsevier
2010 Published
Elsevier Ltd.
All rightsLtd.
reserved
Keywords: Carbon Capture and Storage; Post-combustion Capture; Hazelwood; Integration; Energy Penalty; Cost of Capture
1. Introduction
Coal is the primary fuel for over 80% of Australia’s current power supply and is one of the largest contributors to
Australia's total domestic greenhouse gas emissions. Post-combustion Carbon Capture and Storage (CCS) is seen as
a way of significantly reducing power station emissions however, the challenges in implementing CCS on a large
scale are many. Process Group has designed and built several pilot scale capture plants, including the largest plant
* Corresponding author. Tel.: +61-3-9212-7100 fax: +61-3-39212-7199.
E-mail address: [email protected]
doi:10.1016/j.egypro.2011.02.542
5550
2
T. Dreher et al. / Energy Procedia 4 (2011) 5549–5556
Author name / Energy Procedia 00 (2010) 000–000
on a coal fired power station in Australia. The lessons from these pilot plants combined with a retrofit postcombustion study are presented in this paper. The detailed study looks at carbon capture from a 500MW power
station (nominally 3.7 million tpa CO2) using three different solvents with and without heat integration into the
steam cycle. Given that a large scale pilot carbon capture plant will be built before such a full scale plant, we also
present data pertaining to processing 25% of the flue gas from a 500MW power station (nominally 0.9 million tpa
CO2).
2. Carbon Capture and Sequestration at the Hazelwood Power Station
In early 2009 Process Group completed commissioning of a 15,500 tpa pilot carbon capture and mineral
sequestration plant it designed and fabricated for International Power’s Hazelwood Power Station. Hazelwood is a
1600MW nominal capacity brown coal fired power station that supplies up to 25% of Victoria’s electricity needs,
which is equivalent to approximately 5% of Australia’s National Electricity Market.
The Hazelwood pilot capture plant follows the standard solvent absorption process however, some modifications
to this process were made so that the plant can operate with multiple solvents including BASF’s PuraTreat™ F,
generic amines, and carbonate solvents. A portion of the CO2 captured is sequestered via an innovative process that
results in the production of calcium carbonate and eliminates the need for expensive waste water treatment
chemicals. Waste ash water from the power station contains a high concentration of dissolved calcium hydroxide
and as such has a pH in the order of 12 that needs to be reduced before it can be discharged. The treatment system
designed by Process Group injects the captured CO2 into the ash water where it reacts with the dissolved calcium to
form calcium carbonate. The reaction lowers the solution pH and sequesters the injected CO2 as a solid calcium
carbonate product.
2.1. Issues Facing Small Scale Pilot Plants
Small scale capture plants, such as the unit at Hazelwood face a range of economic challenges that make it very
difficult to demonstrate new technologies, designs, or construction techniques because such advances are frequently
considered too risky to implement. In addition, small scale and demonstration capture pilot plants (<0.3 million tpa
CO2) are of a size that they can be designed and built using standard engineering practices and as such new designs
and technologies aimed at reducing the size and cost of large scale capture plants are often deemed too risky for
incorporation into these demonstration plants. Therefore, most small scale pilot plants do not include technologies,
such as heat integration, alternate vessel construction, or new absorber internals that may have higher technical risk
but offer significant potential for future reduction of the cost of carbon capture.
Funding mechanisms for demonstration CCS projects also often hamper technology development in that they are
funded by research or governmental organisations under a lump sum grant to cover the plant capital (CAPEX) and
operating (OPEX) expenditures over the life of the project. This type of funding has two significant disadvantages;
(i) short project life times discourage the end user to make any significant modifications to existing infrastructure
and (ii) as the funding level is fixed rather than being linked to parameters such as the amount of CO2 avoided or
technology advances, the end user has little incentive to try to decrease OPEX or increase plant efficiency. The
result is that there is little incentive to trial new technology so many small scale pilot plants do little except
demonstrate already proven technology.
3. Large Scale CCS
The International Power pilot capture plant only captures in the order of 0.1% of the Hazelwood Power Station’s
total CO2 emissions. For CCS to significantly reduce emissions from stationary sources much larger capture plants
need to be built. This retrofit study for full scale post-combustion capture from a 500MW power station investigates
the effect of heat integration for three different solvents. This study assumes that certain emerging technologies,
such as the Ramgen Supersonic Compressor, and mechanical limitations, such as fluid distribution in large columns,
will be resolved in the years before such a plant is actually built. While this study should be considered indicative,
all key issues have been addressed and the general conclusions reached are considered applicable to future studies.
T. Dreher et al. / Energy Procedia 4 (2011) 5549–5556
Author name / Energy Procedia 00 (2010) 000–000
5551
3
3.1. Study Methodology
The study examines retrofit post-combustion capture (PCC) on a 500 MW brown coal fired power station in the
Latrobe Valley using three solvents at two separate flow rate cases. It should be noted that due to the low cost and
low sulphur content (<1%) of the coal, such power stations are relatively inefficient and do not have flue gas
desulphurisation (FGD) or nitrogen removal (DeNOx) units. Furthermore coal drying is not installed.
Process simulation of the Power Station steam cycle and its integration into the capture plant closely follows that
described previously for similar studies [1,2]. The economic analysis follows the previously described
CO2CRC/UNSW methodology for CCS projects [3]. In this instance, the actual installed capital costs have been
estimated by Process Group together with standard setup costs and a project contingency. The CAPEX and OPEX
figures quoted are for an installed and fully operational capture and compression plant and include auxiliary costs
such as chemical storage facilities, steam cycle modifications, and costs associated with providing for the increased
cooling water demand. The basis of design is given in Table 1 and although it is for Australian brown coal
conditions the study is sufficiently generic for the conclusions to be relevant for capture plants around the world.
Three solvents were considered for this study; amine, glycinate, and carbonate. It should be noted that amine and
glycinate based solvents are extensively used in current commercial solvent processes however, the carbonate
process presented here is an emerging precipitating solvent technology under development by the CO2CRC.
Precipitating solvent systems have lower circulation rates compared to conventional systems and as such could
reduce plant CAPEX and OPEX. Energy to run the capture plant is obtained parasitically from the power station.
Specifically, all electrical power for the capture plant is supplied directly from the power station’s generator thereby
decreasing the sent out power. Steam supplied directly from the LP turbine is used as the reboiler heat source. In
order to reduce the parasitic energy load, heat integration between the capture plant and the steam cycle was
incorporated into each capture plant design broadly as shown in Figure 1. In this study heat integration refers to
integration of energy from the power station, such as flue gas and condensate from the steam cycle, with the heat
exchangers of both the capture plant and power station. There are many different heat integration options that could
have been investigated and optimised for each specific solvent however, for simplicity and ease of comparison we
have chosen a single heat integrated design. The integrated exchangers were selected so that the condensate is
returned to the deaerator to maintain a deaerator temperature of approximately 170 °C. The design is not optimised
for either the steam cycle nor each solvent and is designed to illustrate the potential efficiency gains that heat
integration offers. It is of note that previous work [2] showed target parasitic loads resulting from heat integration
improvements for a similar power station to be as low as 11-19% depending on the extent to which coal drying was
incorporated into the design. Such low parasitic loads may be possible due to the considerable available heat in the
exhaust flows from relatively inefficient power stations.
3.2. Process Design
The capture design proposed for each solvent largely follows the standard solvent absorption system as given in
Figure 1. The main differences are summarised in Table 2: Due to the low sulphur content of the coal used in this
study FGD or DeNOx treatment was not employed for any solvent however, due to amine’s intolerance to SOx
20wt% caustic scrubbing was used only for the amine solvent. Mechanical vapour recompression was considered
however, for the cases investigated its incorporation was found to be unfavourable as the compressor and associated
equipment significantly increased the plant CAPEX and energy penalty. Nevertheless, such a configuration may
prove to be beneficial in other situations.
In order to investigate the benefits of direct heat integration between the carbon capture plant and the power
station’s steam cycle the following streams were targeted as illustrated in Figure 1: Firstly, the steam required for
the reboiler is desuperheated as it exits the LP turbine and is fully condensed in the reboiler. The condensate is then
pumped through the compressor inter and after coolers, heating it before it returns to the deaerator. Secondly,
condensate from the LP turbine passes through the capture plant condenser and then the flue gas economiser before
it also returns to the deaerator. The flue gas economiser is designed by ERK and pre cools the flue gas before final
cooling in the Scrubber section of the Absorber tower. In the non-heat integrated cases these exchangers are all
cooled via cooling water.
5552
4
T. Dreher et al. / Energy Procedia 4 (2011) 5549–5556
Author name / Energy Procedia 00 (2010) 000–000
Table 1: Design Basis
25% Flue gas processed (125MW equiv.)
Gross/Net power output before CCS
100% Flue gas processed (500ME equiv.
520 / 500 MWe
Thermal efficiency
35% LHV
Flue gas flow rate
809,500 kg/h
Flue gas inlet conditions
3,238,000 kg/h
2 kPag, 192 °C
Flue gas compositions (mol %)
CO2 11.0%, N2 + Ar 62.6%, O2 3.9%, H2O 22.5%
Flue gas impurities (ppmv, dry basis)
NOx 151, CO 13.9, SO2 211, SO3 0.5
CO2 emission before capture
4,949,000 tpa
Overall CO2 recovery rate (mass basis)
85 %
CO2 product (at compressor discharge)
> 99 mol%, 150 bara , 50 °C
Fuel (coal) cost (AUD$/GJ HHV)
0.7
Project life (years)
25
Construction period (years)
2
Plant capacity factor (%)
85 %
Discount rate (% real)
7%
Electricity selling price (AUD$/MWh)
40
Table 2: Process Design Comparison
25% Flue Gas Processed (125 MW equiv.)
Solvent
100% Flue Gas Processed (500MW equiv.)
Amine
Glycinate
Carbonate
Amine
Glycinate
Carbonate
CAPEX (Millions AUD$ 2010 ±20%)
253
253
253
621
615
602
OPEX (Millions AUD$ 2010 ±15%)
23
21
22
64
58
61
LCOE (Millions AUD$ 2010/MWh)
58
58
58
111
108
107
Relative lean solvent flow rate
1.9
2.5
1
1.9
2.5
1
Absorber solvent intercooler
Yes
No
Yes
Yes
No
Yes
Absorber wash section
Yes
No
No
Yes
No
No
Note: All values are indicative and should not be taken as definitive project design. Separate CAPEX, OPEX, and LCOE figures
for the (i) with and (ii) without heat integration cases are not shown however, within study accuracy they are effectively the same
in this instance.
3.3. Plant Comparison
The Cost of Capture
As shown in Table 2, within the accuracy of this study, the plant CAPEX for each of the three solvents is the
same at AUD$253 and $613 million respectively for the 25% and 100% plants. Even though there are some
differences between the solvent designs, this has little effect on the overall plant cost as the capture and compression
plant equipment only accounts for ~50% of the total plant CAPEX. In accurately costed smaller capture plants
(<100,000 tpa CO2) without FGD or DeNOx units Process Group also found very similar CAPEX figures across a
range of commercial solvents further confirming that designing plants to operate with a range of solvents has little
impact on overall CAPEX. As such, this conclusion enables the end user to operate with the best available solvent at
any given time and gives them the ability to take advantage of advances in solvent technology over the life of the
capture plant. The OPEX showed greater variance with a range of AUD$58-64 million, although the difference is
within data accuracy of ±15%. This difference is largely dictated by the cost of replacement and disposal of solvent
5553
T. Dreher et al. / Energy Procedia 4 (2011) 5549–5556
Author name / Energy Procedia 00 (2010) 000–000
5
and other chemicals that can vary greatly between solvents. As such, OPEX figures can vary significantly according
to the solvent chosen and should be determined on a case by case basis.
Figure 2 illustrates that the cost of capture for heat integrated plants, expressed as AUD$ per tonne CO2 avoided,
for the three solvents is in the range AUD$53-63/t CO2 avoided (100% plant). This figure incorporates an
improvement of $8-13/t due to the heat integration. The reduction is largely due to the increase in sent out power, as
discussed below. It is expected that with optimisation of heat integration the cost of capture could be further reduced
to at or below AUD$50/t CO2 avoided for the best option. The results also indicate that the cost of capture is
significantly greater for the smaller scale (25%) plant due to the lower amounts of CO2 captured.
Condensors
Wash water
cooler
Stage 1
Lean solvent
cooler
Stage 2
Intercoolers
Lean/Rich
Exchanger
Regenerator
TEG Dehydration Aftercoolers
CO2 to
pipeline
Solvent
intercooler
Scrubber/
Absorber
Flue gas
economiser
Reboiler
Scrubber
cooler
LP Turbine
HP
Turbine
IP
Turbine
Generator
Deaerator
Desuperheater
Boiler
Figure 1: Schematic of heat integrated capture plant and steam cycle
Energy Penalty
Figure 3 (a) and (b) illustrate that with heat integration the full scale CCS plant results in a 23-28% reduction in
sent out power, which is equivalent to 8-10% reduction in thermal efficiency. Figure 4 shows the impact on cooling
water demand with the inclusion of the capture plant resulting in an 85-95% increase in cooling water duty. Further
optimisation of the heat integration system for this application is expected to decrease parasitic losses to between
15-20%. This would also reduce the cooling water impact, limiting the increase on the base load cooling duty to an
additional 75-80%. Heat integration optimisations could include improved and additional process integration
between the capture plant and the power station such as more complete matching of the available hot and cold
streams, heating boiler feedwater streams downstream of the deaerator, incorporating coal drying options, and using
steam turbine drives for applications such as the CO2 compressor. As discussed above, the improved efficiency
resulting from heat integration significantly decreases the cost of capture.
5554
T. Dreher et al. / Energy Procedia 4 (2011) 5549–5556
Author name / Energy Procedia 00 (2010) 000–000
6
Capture plant cooling water duties vary between the solvents mainly due to differences in the solvent circulation
rate, the condenser operating conditions, and the requirement of a solvent wash section or intercooler however, the
total amount of cooling water required for the capture plant and steam cycle varies less than 15% across all solvents.
Heat integration does reduce the quantity of cooling required in the capture plant however, this reduction is offset by
an increase in demand in the steam cycle due to the increase in sent out power. It should be noted that the basis of
this study was restricted to the use of cooling water and no effort was made to investigate air cooling, which would
allow further reductions in cooling water requirements.
Cost of Capture
$91
$81
$65
$57
$84
$93
$66
$98
$88
$60
$53
$80
$71
$100
$63
AUD$/tonne CO2 avoidedas
$120
100% Without heat integration
100% With heat integration
25% Without heat integration
$40
25% With heat integration
$20
$0
AMINE
GLYCINATE
CARBONATE
Figure 2: Cost of Capture (Bars from left to right 100% with/without heat integration, 25% with/without heat integration)
5%
0%
AMINE
GLYCINATE CARBONATE
10%
10%
8%
8%
4%
2%
2%
2%
6%
3%
2%
7.0%
5.8%
7.2%
6.0%
10%
8.0%
6.8%
20%
8%
10%
11%
10%
12%
3%
2%
25%
15%
Reduction in Thermal Efficiency
(LHV % points)
30%
(b) Energy Penalty - Reduction in thermal
efficiency
14%
28%
23%
35%
29%
24%
40%
32%
28%
Reduction in Sent Out Power
(a) Energy Penalty - Reduction in sent out
power
0%
AMINE
GLYCINATE CARBONATE
Figure 3: Energy Penalty of Capture. Reduction in (a) Sent out power, (b) Thermal efficiency. (Bars from left to right 100% with/without heat
integration, 25% with/without heat integration)
Levelized Cost of Electricity (LCOE)
The LCOE for a capture facility incorporates the added cost and energy impacts of the plant into the most
significant variable for a power generator. The various costs for each of the cases in this study were imposed on a
base plant LCOE of AUD$40/MWh. The calculation of LCOE also includes an allowance of AUD$15/t CO2 for the
storage component, which is considered appropriate for storage locations near to the Latrobe Valley. As given in
T. Dreher et al. / Energy Procedia 4 (2011) 5549–5556
Author name / Energy Procedia 00 (2010) 000–000
5555
7
Table 2 the resulting LCOE ranges from AUD$107–111/MWh for the full scale case and AUD$58/MWh for the
25% case.
90%
86%
89%
85%
100%
98%
120%
95%
100% Without heat integration
80%
100% With heat integration
60%
23%
22%
22%
21%
25% Without heat integration
25%
40%
24%
Increase in cooling water duty
Cooling Water Demand Increase
25% With heat integration
20%
0%
AMINE
GLYCINATE
CARBONATE
Figure 4: Cooling water demand increase of PCC plant (Bars from left to right 100% with/without heat integration, 25% with/without heat
integration)
Scrubber/Absorber Vessel Design
Due to the very large diameter (~22m) and height (~49m) of the Scrubber/Absorber vessel it may be impractical
to construct this vessel from steel. A more practical and possibly cost effective alternative is to construct the vessel
from concrete and line it with thin stainless steel in order to protect the concrete from chemical attack. The
Scrubber/Absorber vessels included in this study were designed to be built in concrete taking into account the loads
from the various internal structures and packing and the relevant environmental conditions. A circular vessel crosssection was deemed most suitable in terms of constructability and cost.
WES Froth Absorber Technology
The Scrubber/Absorber vessel is one of the most expensive equipment items in the capture plant costing in the
order of $20 million (100% plant) and accounts for approximately 3.5% of the total operational capture plant
CAPEX. Given the significant cost of the Scrubber/ Absorber vessel, any means of simplifying the internals or
reducing the absorber height could significantly reduce vessel cost. Revolutionary advances in internals design, such
as the WES Froth Technology proposed by Westec Environmental Solutions [4], could lead to reductions in
Absorber section height by 50% thereby decreasing the Scrubber/Absorber cost in the order of $5-8 million. The
WES technology incorporates novel patented Micro Froth Absorber technology where conventional packing is
replaced by the WES patented froth generators. In this manner mass transfer is facilitated across a froth matrix
generating mass transfer surface areas many times higher than through conventional random or structured packing
thereby resulting in lower absorber heights.
Solvent and Chemical Usage
Solvent and chemical make-up and disposal costs contribute 30-40% to total OPEX however, for generic solvents
this figure would be closer to 20-30% as commercial solvents are at least 50% more expensive than their generic
counterparts. Solvent costs can vary significantly between solvents due to differences in solvent make up rates,
primarily due to differences in solvent degradation rates which are higher for generic solvents, and this factor needs
to be analysed on a case by case basis. For this study we have calculated make up rates based on typical degradation
rates for the given solvent.
The very large solvent and chemical inventories required for full scale capture will present several handling
issues associated with chemical manufacture, transport, storage, and disposal. It is Process Group’s experience that
end users prefer solvents with low toxicity to reduce their risk in case of spills or accidents and to facilitate ease of
disposal. In particular, the formation of carcinogenic nitrosamines and their potential release to the environment may
5556
8
T. Dreher et al. / Energy Procedia 4 (2011) 5549–5556
Author name / Energy Procedia 00 (2010) 000–000
be a major impediment for amine based solvents. Process Group has seen a strong preference by end users for
glycinate rather than amine based solvents due to the elimination of such toxic by-products.
Compression
To render the recovered CO2 vapour in a state suitable for geosequestration, the vapour is compressed and
between compression stages water is removed via conventional CO2 glycol dehydration. The compressor used could
be a conventional (e.g. gear type) multi-stage compressor however, in this study we use a 2-stage Ramgen
Compression Systems supersonic compressor [5] as it offers the opportunity for significant waste heat recovery with
stage discharge temperatures in the order of 250 °C. Compared to conventional technologies Ramgen’s shock
compression technology represents a significant advancement in the state of the art for many compressor
applications and specifically for CO2 compression. The principle advantage of Ramgen’s shock compression is that
it can achieve high compression efficiency at very high single stage compression ratios resulting in a product
simplicity and size that will lower both manufacturing and operating costs. Costing ~$28 million (100% case) the
Ramgen compressor accounts for ~4.8% of the total plant CAPEX. The compressor also accounts for ~75% of the
CCS plant’s total power usage to drive the compressor motors. For the three solvents the 100% case requires 1013% of the power station’s total generation capacity to run the compressor. Since a greater amount of power is
required as the suction pressure decreases, high regeneration pressures favour lower compressor power usage.
4. Conclusion
The following conclusions are drawn from this study and previous work performed by Process Group on small
scale capture plants without FGD or DeNOx facilities:
x There is <5% difference between existing solvent processes in terms of overall plant CAPEX and <15%
difference in OPEX.
x The most advantageous way to design a capture plant for the end user is to make it flexible so that it
functions with a number of commercial solvents and that it is able to be adapted to enable the end user
to change the solvent or operating conditions to take advantage of advances in solvent design.
x The cost of post-combustion capture for a retrofit to a 500 MW brown coal power station with nonoptimised heat integration is in the range AUD$53-63/t CO2 avoided. This figure incorporates an
improvement of approximately $8-13/t due to the heat integration. With further design and process
optimisation performed specific to the given power station and capture plant it is expected that the cost
of capture could be reduced to at or below AUD$50/t CO2 avoided.
x A primary concern of the end user is chemical safety and as such there is a strong preference for
solvents that exhibit low toxicity and low volatility.
5. Acknowledgements
The authors gratefully acknowledge the cooperation of the following organisations in this study; CBI
Constructors, Cooperative Research Centre for Greenhouse Gas Technologies (CO2CRC), ERK Eckrohrkessel,
International Power, Leighton Contractors, and Ramgen Compression Systems. Thanks are also extended to Dr. M.
Ho (CO2CRC/UNSW) for assistance with the economic analysis.
6. References
[1] Harkin, T., Hoadley, A, and Hooper, B., Process integration analysis of a brown coal-fired power station with
CO2 capture and storage of lignite drying, Energy Procedia 1, 2009, 3817-3825.
[2] Harkin, T., A. Hoadley, and B. Hooper, Reducing the energy penalty of CO2 capture and compression using
pinch analysis, Journal of Cleaner Production, 2010, 18(9), 857-866.
[3] Allinson, W.G., Neal, P.R., Ho, M., Wiley, D.E. and McKee, G.A., 2006. CCS economics methodology and
assumptions, School of Petroleum Engineering, The University of New South Wales, Sydney, Australia. CO2CRC
Report Number RPT06-0080.
[4] Westec Environmental Solutions (WES), www.wes-worldwide.com
[5] Ramgen Compression Systems, www.ramgen.com